Keith Hopkins possessed many gifts as a scholar. He had a rare ability to ask good questions. He had a knack for seeing the big picture and imagining the way a puzzle, with a lot of missing pieces, might have fitted together. His broad learning gave him comparative perspective. He had a disarming and even seductive style, with a panache for playing up how speculative his own argument was. His scholarship was ranging and generative, and perhaps, depending on one's tastes, a little blithe. Surely no one ever accused H. of being so patient that he bore down on a problem for too long. H. was certainly sui generis.

Fittingly, the true medium of H.’s work was the essay. H.’s only proper monograph, A World Full of Gods (2000), received a mixed reception at best, and it cannot be said to have made any lasting impact. And yet, in stark contrast, H. wrote at least a dozen shorter studies — articles, conference papers — that are true classics. In at least a few cases (e.g. the famous ‘Taxes and trade’ article from this journal in 1980), it is hard to imagine the trajectory of the field without his intervention. So it is a genuine service that Christopher Kelly has brought together, almost a decade and a half after H.’s death, fourteen of H.’s previously published works, along with a dozen brief Afterwords from other scholars offering commentary and assessment. Kelly's lengthy introduction (54 pages) opens the volume and provides a biographical survey and intellectual appraisal. Probably anyone except those very close to H. will find some fresh information that gives a rounder picture of the man and the scholar. Kelly also sets the tone for the volume as a whole — generously appreciative but not quite panegyrical. A reviewer should not spoil all the juicy bits, but the reader can expect to find ‘slapdash’, ‘downright lazy’ and ‘irritating’ thrown around by the commenters, though none of the Afterwords are impious. This book is through and through a Cambridge production, a treatment of the late lamented H. by the home team. But that is to be expected, and it does not detract from the value or interest of this collection, which needed to be brought together to have a worthy memorial of one of the great social historians of the ancient world.

It would be impossible, and tedious, to summarise each of the chapters, so let a few general observations suffice. H. made major original contributions to the study of Roman demography, society, economy, power and religion, all represented in this collection. H. was both a classicist and a historical sociologist. He was a bold model-builder, willing to make big assumptions and then to test their implications against the inevitably limited evidence in the possession of the ancient historian. What allowed H. to be so successful as a model-builder was not his command of the ancient sources or finesse in the interpretation of ancient texts or documents (though these could be formidable); in one sense, quite the opposite. His strengths were his wide interests and familiarity with other pre-modern societies. These put limits to the plausible and suggested how the fragments might be reassembled into an internally consistent whole. The comparative approach, for instance, underlies his contributions to Roman demography. While H. was willing to allow that Rome could be exceptional (e.g. in the size of the imperial capital, or the extent of slavery), he knew that such claims had to be made carefully and thought through. In other words, if Rome was the biggest city in pre-modern Western history, then how was that accomplished? If the Romans had one of history's few ‘genuine slave societies’, what were the structural implications of that fact?

H.’s approach was eclectic and undogmatic. His interventions in the study of the Roman economy started to break down the primitivist/modernist dichotomy. The influence of Finley was strong, yet H. was anything but a rigid partisan. He was not the first to ‘measure’ the Roman economy, but his quantitative forays were more provocative and more systematic in their imaginative architecture than anything that had come before. And if H. did not drill down on any one problem for too long, he returned to a few main themes over and over, and was willing to change his mind. One of the few regrettable omissions in this collection is that we are given the original 1980 ‘Taxes and trade’ paper (ch. 6) but not the later 1995/6 and 2000 versions of the argument, which deserve to be read together to follow the development of his thought.

H.’s methods were fertilised by a wide range of influences, and while he is best remembered as a model-builder, the second main tendency of his approach was a daring historical empathy that tried to imagine what it was like to experience the Roman Empire. His experiments in this vein are a sort of historical anthropology. Some of them are still astonishingly stimulating, for instance the essay on Roman slavery as we might reconstruct it from the biography of Aesop (ch. 11), reprinted here with an appreciative Afterword by Catharine Edwards. There may not be a more valiant attempt to do the impossible, which is to enter the emotional texture of the master-slave relationship in the Roman world. Yet other forays along these lines fall flat, such as his fabricated letter from Heaven by a dead Septimius Severus, with some ornery fake correspondence between H. and thinly disguised modern scholars (ch. 14). The jabs at Syme make for amusing reading, but are not particularly edifying; Mary Beard's Afterword to this piece manages to give an honest judgement without being unkind.

H. was almost unmatched in his ability to imagine how the Roman Empire worked as an interlocking system of power. His essay on ‘The political economy of the Roman Empire’ puts these talents on display (ch. 13). It was edited by Walter Scheidel and published after H.’s death in a 2009 collected volume. There is not a more stimulating summary of how the Roman system worked. It is a masterpiece on a par with any of H.’s work and the ultimate vindication of his sweeping vision over the dry prosopographical approach so much in favour during H.’s formative years. It deserves to be better known, and perhaps the essay's inclusion here will help it gain the reception it deserves.

H.’s most important contributions, spanning a lifetime and now standing together between two covers, present us with a central paradox. H. could not have built his models without the patient and painstaking labours of the scholars who reconstructed the individual bricks of the edifice. But H. was critical of the tendency in the field to look only at the bricks themselves, without imagining what the edifice as a whole looked like. A favourite question of H. was ‘So what?’. Reading these essays reminds us that it is still the best question.